Epistemic norms 2010

1

Epistemic norms

In Sven Berneker and Duncan Pritchard eds The Routledge Companion to Epistemology , 2010

1. Epistemic normativity When we evaluate our beliefs and our reasonings as justified or unjustified, as good or bad, or as rational or irrational, we make, in a broad sense of the term, normative judgements about them. It is often said that justification itself is a normative notion, and that epistemology is a normative discipline. But the nature of epistemic normativity is elusive, and there are several strands in these debates, many of which are common to the domain of practical or ethical normativity and to the domain of epistemic or cognitive normativity. (i) Norms and values. It is common to distinguish two kinds of normative notions, those which are deontic, formulated in terms of oughts, right or wrong, prescriptions and permissions, and those which are teleological or axiological, formulated in terms of good, bad, valuable, virtuous or defective. There are important differences between these two kinds of normative properties. The former in general call for actions, can give rise to sanctions and do not admit degrees, whereas the latter are more a matter of possession by the agent of a certain kind of sensitivity, call for praise or blame and are comparative. Deontic notions are often associated to “thin” properties, whereas axiological ones are often “thick” properties. One main problem about epistemic normativity is whether it should be formulated in terms of epistemic norms in the deontic sense or in terms of value-based notions and whether one has priority over the other. (ii) Norms, rationality and reasons. The notion of norm is commonly associated with the notion of rationality, both because norms concern what one ideally ought to do or to think and because rationality seems to be normative in the sense that it prescribes a certain kind of conduct. But it is not clear in what sense rationality is normative rather than descriptive of agents or believers. There is an important sense in which what rationality requires normatively differs what one has a reason to do or to think. Some think the latter has priority over the former. The “buck passing” account of values in the ethical domain (Scanlon 1998, Skorupski 2010) according to which all normative concepts (including the teleological ones) should be formulated in terms of reasons also concerns epistemic reasons and whether these have priorities over values. (iii) Normative regulation. A normative statement is usually supposed to govern or to guide the actions of those who are subject to it. What kind of guidance or regulation does a norm imply? Mere permissions and

2

recommendations, or strict prescriptions? And to what extent is the agent who obeys a norm supposed to be aware that he is subject to it? To what extent does ought imply can and if it does what is the nature of the normative can? This depends in part on whether the normative ought is considered as categorical (prescribing regardless of any goal) or hypothetical (relative to a goal for which a prescription is instrumental). (iv) Normative objectivity. Are norms more like rules or conventions, which depend upon human decisions, or more like universal principles governing the whole domain of thought and action? Are they necessarily plural or can they be unified or ranked along a hierarchy? Are there some basic norms which rule a whole domain? This is closely related to the question of the objectivity of norms and to their ontological status: are there genuine normative facts and properties which make our normative statements true or false or are normative statements mere expressions of our attitudes? Should one be a cognitivist (Wedgwood 2007) or an expressivist (Gibbard 2008) about them? Epistemology itself can be understood as a meta-normative discipline: it not only deals with epistemic norms and values, but it also has, at a meta-level, to evaluate the appropriateness and objectivity of these norms. Are there standards to evaluate our standards, meta-norms about which norms we should adopt? And if there are such metanorms, how to choose between them? If epistemic norms and values are objective, some must be such that they cannot be evaluated further. (v) Naturalisation. Do norms supervene on natural facts? Can they be reduced to these? Unless one adopts a form of non factualism (Field 2009) or a form of eliminativism (Weinberg, Nichols and Stich 2001), to the effect that the apparent objectivity of normative discourse is illusory and that norms can be reduced to descriptive facts, one has to address the question of the relationship between the normative and the descriptive. The epistemological counterpart of this question concerns how much empirical matters count in epistemology and how much it can depend upon psychology or biology (Kornblith 1982). Questions (i)-(v) all arise for the practical or ethical domain and one can expect that they can be transposed to the epistemic domain. But there are important asymmetries. In particular there are epistemic norms in the deontic sense, and if they entail specific duties to believe, it is not clear that what we ought to believe or not to believe must issue in actions, for believings are not, in general, things that we do. Moreover, if there are epistemic obligations or duties about beliefs, we must be able to conform to them and to perform certain kinds of actions (ought normally implies can); but if, as many philosophers think, beliefs are not actions and are not under the control of the will, we cannot conform to such obligations, and their existence is dubious. So it is not clear that responsibility for belief, or the capacity to be subject to praise or blame in the epistemic domain, must rest upon the existence of epistemic duties. These questions lie at the heart of the “ethics of belief” debate (Clifford 1879, James

3

1905, Feldman 2000). Another reason to deny the parallelism between the practical and the epistemic case is that although the idea of a plurality of ethical values, reasons or norms makes sense in the former, it does not make sense in the latter: there may be many kinds of good reasons for which one does something, but normally a belief is justified only by one kind of reason - its being true or its being supported by evidence. One can believe for practical, rather than for epistemic reasons. But the former are, in comparison to the latter, the wrong kind of reasons. Wishful thinking is a normatively incorrect kind of believing, although it can be correct in the practical domain (if I am indifferent in my choice between A and B, I can choose either, but if I have equal evidence for believing p or q, I must suspend judgment). In spite of important structural analogies between practical and theoretical reasoning (Audi 2000, Kelly 2003), there are important disanalogies. Pragmatism is in general the view that practical values and norms can be compared to – and sometimes can overrideepistemic norms. But if the normative landscape differs from the practical to the theoretical realm, pragmatism is less plausible. Because they do not give the same kind of answers to questions (i)-(v), not all epistemologists believe in the existence of epistemic norms. Some think that they reduce to epistemic values and favour a teleological or value based account of epistemic normativity. Others reduce it to the normativity of rationality or to the normativity of reasons. Yet others reject the very idea of normative authority in the epistemic domain, because they consider that epistemology has to be naturalised, although they do not reject the existence of epistemic values altogether.

2. Norms of rationality If there are epistemic norms, what are they? One can understand these, in the first place, as general requirements of rationality flowing from the very nature of belief and of the mental. For instance two main requirements of rationality are those of coherence and deductive closure: (a) one ought not to believe p and not p (b) one ought to believe that q if one believes that p and that p entails q But what is the nature of this ought? One can take it as a constraint on rational belief or as a general principle for the interpretation of any believer as a rational being (Davidson 1980, Millar 2006). In this sense the principles of logic, or those of probability theory, are the most general norms for belief, and no subject can be said to have a belief unless these norms can be applied (Davidson 2004). The problem with such very high profile norms is that they do not seem to be

4

normative at all. For a principle to be genuinely normative, it must have normative force and be able to actually regulate belief. It must also have normative freedom, in the sense that one must have the possibility of violating it. But general or abstract norms of rationality like (a) and (b) have neither of these properties, since they are idealised descriptions of a rational believer: at best they tell us what the constitutive properties of rational belief are, but they offer us no guidance (Railton 2000). Perhaps normative statements like (a) or (b) apply to what it is to have a mind or being interpretable as a thinking being, but they do not prescribe (Schröder 2003, Engel 2008) Nevertheless ideals of reason or normative requirements like (a) and (b) can be violated: one can certainly have contradictory beliefs or fail to draw a conclusion from one’s other beliefs. Moreover, one can appreciate the force of the ideal and yet fail to conform to it. If one believes that p, and believes that p entails q, one can fail to infer q, if one has a reason not to believe q. For instance if you believe that the earth was created in less than a week, and believe that if the earth was created in less than a week, you are rationally required, or committed to believing that the earth was created in less than a week, but certainly you ought not to believe this, since it is false, and you have good reason or justification to believe that it is false. The difference between this latter ought and the one which figures in the rational requirement (b) is often expressed in terms of the idea that rational requirements take wide scope: (a’) You ought (if you believe that p, and that p implies p, believe p) From (a’) one cannot detach the conclusion ‘q’, whereas inferences like the above about the origin of the earth take narrow scope: (a’’) If you believe that p, and that if p then q, then you ought to believe that q which is invalid. On one conception (Broome 2000) what it takes to be rational always involves wide scope normative requirements, on another (Kolodny 2005) to be rational always implies having a reason and rationality is only a matter of narrow scope. Non detachment is a symptom of the difficulty of applying the ideals of reason to particular cases. We need to know how ideal rational norms like (a) and (b) regulate the epistemic attitudes of agents. Because we are not frictionless beings and can be irrational sometimes, some conclude that rational requirements, in so far as they express the norms of logic, are impotent and that logic has no relevance to actual psychological reasoning (Harman 1986). But still, one can learn logic, and it can improve our reasoning ability, just as one can learn epistemic rules. The problem is how to implement them so that they actually guide our epistemic practices.

5

3. Epistemic norms and epistemic concepts Epistemic norms can be formulated as correctness conditions relating a kind of epistemic state with an objective condition (Mulligan 2008), and the relation between these is plausibly understood as a necessary and sufficient condition: S judges (believes) correctly that p iff the proposition that p is true conjectures correctly that p iff it is probable that p S has a correct interrogative attitude towards p iff it is questionable whether p S doubts correctly whether p iff it is doubtful whether p S is correctly certain that p iff if is certain that p Alternatively we may think of them as “possession conditions” for various kinds of concepts (Peacocke 1992), which are typically associated to the kind of epistemic justification which they provide. Thus perceptual concepts obey different kinds of “normative liaisons” from judgmental concepts or from logical concepts. But such a scheme is bound to be abstract unless one tries to spell out the relation between the correctness condition and the believer (doubter, interrogator). The more plausible proposal is that epistemic norms are conditions of “epistemic permissibility” for beliefs, which issue mays rather than oughts depending on the kind of epistemic state (Pollock and Cruz 1999): (c) If it perceptually seems to you that p, then you are prima facie rationally permitted to believe that p (perception) (d) if you are permitted to believe that p and that if p then q, then you are prima facie rationally permitted to believe q (deduction) (e) if you have observed n As and they have been discovered to be Gs, then you are prima facie rationally permitted to believe q (induction) (f) if someone testifies to you that p, then you are permitted prima facie rationally permitted to believe p Such norms of permission are relative to each kind of epistemic domain (reasoning, perception, testimony, and so forth), they spell out specific epistemic rights or permissions to believe, which are rational and prima facie. Let us suppose that they are correctness norms. These raise several kinds of questions. First what is the nature of the kind of justification that these norms codify? For (c) at least it is not clear that it is a kind of propositional justification, licensed by a transition from beliefs to belief, since the inputs of perceptual beliefs need not, at least on many views of perception, trade in propositional contents. The relation of prima facie rational permission is often called “entitlement” (Burge 1993, Peacocke 2004), and it is meant to capture a kind of

6

default justification. But the exact nature of this entitlement is a matter of controversy. Second, if we understand (c)-(f) (but (a)-(b) as well) as rules specifying the kinds of circumstances in which an agent is licensed to believe, how do they regulate? If we understand them on the model of rules (such as rules of chess), we can construe these in two ways (Boghossian 2008): as a conditional proposition specify a permission (or an obligation) of the form If condition C obtains, you may (must)  or (for requirements of the form (a) or (b) ) as an imperative of the form If C, do  ! The imperative construal does not seem appropriate, since a mere imperative does not wear its normative status on its sleeves: it is only an instruction, and it does not tell us what kind of normative requirement is at issue (permission or obligation? of a practical or of an epistemic kind?). The propositional reading is better in this respect, but it seems overly intellectualistic: in order to obey a norm do we have to believe a proposition which expresses it (one follows, as Wittgenstein says, the rule “blindly”)? In addition it faces easily the kind of regress made famous by Lewis Carroll’s tortoise (Carroll 1895): for if I need a second propositional norm to obey a norm, I need another to obey the second, etc. Perhaps, as Pollock and Cruz (1999) suggest, the kind of knowledge that we have of epistemic norms is procedural, rather than declarative knowledge of a propositional form. A further problem which the present proposal faces is that of its common association with the deontological conception of epistemic justification (Chisholm 1977, Steup 2001, Alston 1989). It seems to imply it: a belief is justified if and only if it held in accordance with correct epistemic norms, and if it respects epistemic obligations. But is not clear that the existence of epistemic norms entails the deontological conception of justification. The norms could be necessary although insufficient conditions for justification. And in the entitlement sense, they typically do not entail obligations or oughts. For similar reasons, the notion of epistemic norm is often associated to a kind of internalism in epistemology, and to the idea that the subject must have an internal access to his reasons or justifications. But in so far as the norms of kind (a)-(f) are concerned this needs not be the case. Perceptual norms of kind (c) and the entitlement to which they give rise need not, and generally do not, entail the existence of a conscious access of the agent to them. Moreover the norms can be understood in externalist fashion of the general form If p is generated by a reliable cognitive process it is permissible to believe p

7

(perhaps along the lines of the “J-rules” of justification proposed by Goldman 1986), where the agent needs not know about the existence of the reliable process. Nevertheless it becomes hard to say, in this externalist construal of rules and norms, that they can act as reasons to believe.

4. The Ur-norm of truth Are epistemic norms diverse, in the way rules, maxims, recipes, or heuristics can be? Even when we conceive them as rules for inquiry, such as Descartes’ rules for the direction of the mind, they have a certain systematicity (Descartes’ rules are guided by a certain conception of intuition as the basic source of knowledge). But can’t we say that there is a hierarchy and that some norms enjoy more prominence than others? These, in order to have the proper kind of generality, have to be categorical and expressed in terms of oughts rather than in terms of permissions or entitlements, and which we can formulate as above in the manner of correctness conditions. There are three obvious candidates. The first, along the lines of (a) above is the rationality norm: (NR) For any p, a belief that p is correct if and only if it is rational or in categorical terms: (NR0) One ought to be rational in one’s beliefs The second one is the norm that evidentialists like Clifford (1878) likes so much, that we ought to believe on the basis of sufficient evidence: (NE) A belief is correct if and only if it is based on appropriate evidence None of these are unproblematic. There are several criteria of rationality. The minimal one is logical coherence or non contradiction, which can also, if one admits degrees of belief, expressed through some form of probabilistic coherence. There are also several kinds of concepts of evidence, and the amount of “sufficient” evidence that a subject needs to have is left undetermined by the statement of (NE) (Owens 2000). A more obvious candidate for the role of general norm for beliefs is the socalled norm of truth for beliefs (Wedgwood 2007, Boghossian 2003, Shah 2003, Engel 2007) which is usually considered as the best expression of the familiar metaphor that belief has an “aim”, which is truth:

8

(NT) For any p, a belief that p is correct if and only if p is true It is natural to interpret this in terms of an ought rather than in terms of a may statement: (NTO) For any p, one ought to believe that p if and only if p which, given that it is a biconditional can be broken into two parts : (NTOa) For any p, if p one ought to believe that p (NTOb) For any p, one ought to believe that p only if p which respectively seem to capture what James (1905) calls “the two ways of looking at our duty in matter of opinion: knowing the truth and avoiding error” (NTB) seems more fundamental than (NR) and (NE). For it seems that when we look for rational beliefs or for belief supported by evidence, it is because these properties enhance their chance of being true. A set of beliefs which would be rational but false would be of little help to an inquirer, and evidence is evidence for truth. In addition, if belief plausibly is taken, among all propositional attitudes, as the central one (Boghossian 2008), (NT) seems to enjoy a kind of priority. But even if we grant that (NT) is the Ur-Norm, (NTB) and (NT) in both formulations raise several problems. In the first place, (NT), like the correctness conditions set up at the beginning of § 3 above, seems more or less trivial. Does it say more that to believe that p is to believe that p is true? If so how can it have normative force and a power of prescription? So the objection that a mere analytic principle of rationality cannot create a genuine norm reproduces. In the second place (NTO) seems clearly wrong. It says that for any proposition which is true, one ought to believe it. But certainly no one is under the obligation, if only under the rational requirement, of believing any true proposition whatsoever. There are too many trivial, uninteresting, or useless propositions which no one cares to believe. In the third place, even if one takes (NTO) in the second , more plausible (NTOb) , which says that if one ought to believe that p, then p is true, nothing seems to follow, since the first member of the conditional can be either true or false; if it is true, nothing follows about what the subject ought to believe, and if it is false, it just says that the subject lack an obligation to believing something, which is not the same as a positive obligation to believe (Hattiangadi 2005). These difficulties illustrate the regulation problem: how can the mere (propositional) statement of a norm wear on its face, as it were, the way in

9

which it should be obeyed? If (NT) is understood in the intellectualist sense as prescribing, on the part of every believer, an intention or goal to accept a belief if and only if it reaches the truth (Velleman 2000), the proposal is too intellectualistic to be convincing: not only the norm (NT) need not be explicitly present to the mind of the believer, but it need not be present at all, even tacitly, for there are too many cases where one believes that p without believing it for the sake of truth, but of a desire for it to be true. A more promising solution to the regulation problem exploits one important feature of ascriptions of belief (Moran 2000, Shah 2003): when one deliberates about whether to believe that p, the best way to do so consists in asking oneself whether p, and a positive answer settles the question. In other words, belief is “transparent” to the truth of the proposition believed. This holds only for conscious, deliberate belief, but it also seems to belong to the very concept of belief. Possessing the concept of belief and understanding what it means to possess it entail obeying what the norm of truth prescribes. This feature may bridge the gap between the abstract statement of the norm (NT) and its realisation within a believer’s psychology, for a belief may be regulated for truth without being explicitly attended in a deliberation about whether to believe that p (Shah and Velleman 2005, Engel 2007). 5. The evidential norm A more common objection to the truth norm, which is implicit in most attempts to resist the idea that it can have normative force, is based on the suspicion that it presupposes a form of evidentialism, the view according to which one ought to believe that p only for evidential reasons, pertaining to the kind of evidence that one has for p’s being true. There are a number of formulations of the evidentialist norm, including (NE) above, Clifford’s maxim, and its explicit association with a certain theory of justification according to which a belief that p is epistemically justified if and only if it fits the evidence that one has of p at a given moment (Conee and Feldman 2005). Whatever its proper formulation can be, evidentialism faces, at least since Pascal’s famous wager argument and James’ will to believe objection to Clifford, the following challenge: given that our reasons to believe - at least in the sense of our motivating reasons - fall short of being our exclusive reasons, it seems perfectly correct to believe that p for (normative) reasons other than truth and evidence, in particular pragmatic or prudential reasons. In this sense, certain kinds of belief behaviours described currently as self-deceptive or irrational can be rational in the practical sense. A wife’s observation of lipstick on a husband’s collar is an evidential reason for her to believe that the husband is unfaithful but it can coexist with a practical reason to believe that - contrary to evidence - the husband is faithful. More importantly there can be a possibility of weighing epistemic and practical considerations for believing, for instance if the benefit of

10

a not too well evidentially supported belief (say that one will recover from cancer) outweighs the benefit of believing the contrary (that one’s life is threatened). On this view epistemic reasons to believe are only contingently normative, and they can be overriden by reasons to desire to believe; hence the so-called constitutive norms of truth (NT) or of evidence (NE) are not constitutive at all. Where the evidentialist claims that the only permissible attitude towards a proposition p is the attitude that one has towards its truth or its evidence (namely belief that p), the pragmatist claims that the contrary attitude (belief that not p) or another attitude (say, accepting that not p, or taking a certain prudential stance towards it) is licensed, if a certain goal of the agent favours its being rational (Foley 1993). If so the permissibility norms above become indeed very permissive, to the point of being only relative and purely contextual (White 2005). The trouble with this line of thought, however, is that it is hard to conciliate not only with the existence of asymmetries between epistemic reasons and practical reasons mentioned above (§ 1), but also that it implies, most implausibly, that many prudential or pragmatic beliefs are under the kind of “manipulative” or “managerial” control which a subject can have when he manages to cause himself or herself to believe something for the sake of a certain practical goal ( as, for instance when, knowing that I am prone to be late and forgetful about the time of an important meeting, I set my watch 10 minutes ahead of the actual time, in order to create in myself, when the time arrives and when I look at my watch, the belief that my meeting is forthcoming). But such a kind of managerial control is necessarily indirect, and unable to answer the constitutive reasons for belief by definition. One way or another, pragmatic reasons to believe seem perfectly extrinsic, and the idea that belief has an aim cannot simply be reduced to the idea that it has one purpose or other (Owens 2003, Ryan 2003). The pragmatist, in order to defend the view that there is nothing intrinsic in epistemic norms and reasons for believing, needs to deny the transparency feature of the regulation of belief (Shah 2003, Hieronymi 2005).

6. Epistemic norms and epistemic values The pragmatist’s challenge to the truth and to the evidentialist norms is the strongest form of a worry that epistemic norms might not trade in oughts, but in goals, or in ends, and that they owe their force not to some categorical imperative or even permission, but to the existence of certain epistemic objectives. The most extreme view of this sort would simply be a kind of relativism, according to which there is nothing more in epistemic norms that certain kinds of policies, which one may adopt in one circumstance required by inquiry, but that one may withdraw in another circumstance (Field 2009). A less relativistic view would claim that the oughts of thought and reason are not as

11

general and context independent as rationality principles or Ur-epistemic norms, but can be, in some sense, specialised. Some oughts may relate to a kind of function that a given role enjoys, like professional oughts: just as the role of a teacher is to teach or of a lawyer is to defend his clients, the oughts of belief might be so specialised (Feldman 2000). And given that there is an obvious connexion between normativity and the performance of a function, which has been exploited by many philosophers who hope to reduce norms to biological functions (Millikan 1996 Papineau 2003) it may promise a short way to naturalistic reduction. In a different spirit some theorists hold that norms are related to social functions, or to various commitments implicit in human discourse, which are ultimately social (Brandom 1994, Chrisman 2008). But going too much in that direction is unpromising if it is supposed to explain epistemic normativity in expressivist terms. We have a sense, which the cognitivist conception of normativity captures better (Wedgwood 2007, Cuneo 2008), that the oughts of belief are more general, and that what thought requires outpasses. One way to defend this view consists in reflecting upon the intimate relation that belief has to knowledge. Whether or not one accepts the strong kind of externalism according to which knowledge is a primitive mental state distinct from belief, it makes sense to suggest that the aim of belief is knowledge, and that it is knowledge which gives to epistemic norms their intrinsicness (Williamson 2000). In this sense, it may be that the norm of belief, both in the constitutive sense and in the regulative sense, is a norm for knowledge This still does not solve the ontological issue associated to question (i) and (v) of § 1 above. The generality of epistemic norms can be captured through the idea that they are, in some sense, related to values, and one may think that the truth and the evidentialist norm can be translated into the idea that the ultimate aim of belief is truth, because truth is the ultimate cognitive value (Lynch 2005). Although the pragmatist will here agree that the normative concepts are better captured in teleological terms, this needs not detract from the irreducibility of the truth goal. Space is then left to understand the normativity of epistemic norms in terms of this goal, and to understand normativity as related to features of our capacities to reach “aptly” the truth goal (Sosa 2007), whatever it is, or to understand it, in a more full-blooded Aristotelian sense, as the achievements of a virtuous agent, in the way a virtue epistemology does (Zagzebski 1996, Hookway 2000). Space if left too to the debate between the monist about the cognitive goal, who believes that epistemic achievement is related to only one kind of objective, truth, and the pluralist who allows others goals, like understanding (Kvanvig 2003). There are indeed a number of conceptions of epistemic value, which are all candidates at accounting epistemic normativity in teleological terms (Pritchard 2007, Haddock, Millar and Pritchard 2010, Grimm 2009). But whether or not these teleological accounts are correct, they will have to make sense of two facts: that the evidentialist norms are, in some sense which is

12

yet to be spelled out but which all friends of the idea of an epistemic norm grant, constitutive of the epistemic domain, and that they can guide or regulate inquiry. But it is not clear that the constitutivity claim amounts to the same as the goal claim, and that they can, in either way, be reduced to each other. For it is perfectly possible to accept that the truth and evidentialist norms set general conditions on belief in general, whereas the considerations pertaining to the value of a given belief in various circumstances pertain to inquiry, or the process of formation and acquisition of beliefs. In order to understand the concept of a belief and of an epistemic state in general, one needs to understand the first, but the second, which are relevant to inquiry, need not coincide with the constitutive norms for belief and knowledge. On this view there would be two different kinds of epistemic norms: those which are attached to a believer’s understanding of the very concept of belief on the one hand, and those which apply to the exercise of belief in inquiry, on the other hand. The norms for belief apply categorically to any believer, whereas the inquirer attends to a goal and to the complex ways of achieving it. The former may be more normative in the constitutive sense, whereas the latter may be more instrumental and teleological. But it has yet to be spelled out in what sense they are related.

References Alston, W.P. 1989 Epistemic justification, Ithaca, NY, Cornell University Press Audi, R. 2000 The Architecture of Reason, Oxford : Oxford University Press Boghossian P. 2008 “Epistemic rules”, Journal of Philosophy, vol. 105, no 9 pp. 472-500 2008 a Content and justification, Oxford, Oxford University Press Brandom, R. 1994 Making it Explicit, Harvard: Harvard University Press Broome, J . 2000 “Normative requirements”, in Dancy ed 2000, pp. 78–99. Burge, T. 1993 “ Content Preservation”, Philosophical Review, Vol. 102, No. 4., pp. 457-488 Carroll 1895 “What the Tortoise said to Achilles, Mind n.s., 4, pp. 278–80 Chisholm, R. 1977 Theory of knowledge , 2nd ed., Prentice Hall Clifford, W.K. 1879 “The Ethics of Belief”, in Lectures and Essays, London : Mc Millan Chrisman, M. 2008 “Ought to Believe” , Journal of philosophy 105, 7, pp 346-370 Conee E. Feldman , R. 2005 Evidentialism, Oxford: Oxford university Press Cuneo, T. 2008 The Normative web , Oxford: Oxford University Press Dancy, J. ed 2000 Normativity, Blackwell: Oxford Davidson D. 1980 Essay on Actions and Events, Oxford: Oxford University Press. 2004 Problems of Rationality, Oxford: Oxford University Press Engel , P. 2007“Belief and Normativity”, Disputatio , II, 23, 179-205 2008 “Davidson on Epistemic Norms » in C. Amoretti & N. Vassalo eds, Davidson,

13 Truth, Language and Interpretation, Munchen: Ontos Verlag, 123-146 Feldman, R. 2000 “The Ethics of Belief” Philosophy and Phenomenological Research, Vol. 60, , pp. 667-695, repr in conee and Feldman 2005 Field .H. 2009 “Epistemology without Metaphysics”, Philosophical Studies, 143, 249-90 Foley, R. 1993 Working Without a Net , Oxford: Oxford University Press Gibbard, A . (2008). “Rational credence and the value of truth”. In T. Gendler & J. Hawthorne (Eds.),Oxford Studies in Epistemology 3. Oxford: Oxford University Press. Goldman, A. 1986 Epistemology and cognition, Harvard: Harvard University Press Grimm, S. “Epistemic Normativity” , in Haddock, Millar and Pritchard 2009 Haddock, A. Millar, A. and Pritchard, D. eds 2010 Epistemic Value, Oxford: Oxford University Press Harman, G. 1986 change in View, Cambridge Mass: MIT Press Hattiangadi, A. 2005 Oughts and Thoughts, Oxford: Oxford University Press Hieronymi, P. 2005 "The Wrong Kind of Reason," The Journal of Philosophy 102, no. 9, 437–57. Hookway, C. 2000 “Epistemic Norms and Theoretical Deliberation', in Dancy ed 2000, 60-77 James, W. 1905 ‘The Will to Believe’, in id., The Will to Believe and Other Essays in Popular Philosophy, and Human Immortality (New York: Dover, 1956), 1–31 Kelly T. 2003 “Epistemic rationality and instrumental rationality: a critique”,Philosophy and Phenomenological Research vol.66, no.3 May 2003, pp.612-640. Kolodny N 2005 “Why be Rational?”, Mind 114(455):509-563 Kornblith, H. 1982 “Epistemic Normativity”, Synthese, 94, pp. 357-376 Kvanvig , J 2003 The Value of Knowledge and the Pursuit of Understanding, Cambridge, Cambridge University Press Lynch, M. 2005 The value of Truth, Cambridge Mass, MIT Press Millar, A. 2006 Understanding People, Oxford: Oxford University Press Millikan, R.G. 1996 White Queen Psychology, Cambridge, Mass, MIT Press Moran, R. 2000 Authority and Estrangement Princeton: Princeton University Press Mulligan, K.2008 “Intentionality, Knowledge and Formal Objects”, Disputatio, II, 23, 205228 Owens , D. 2000 Reason without Freedom , Routledge 2003 “Does Belief has an Aim?”, Philosophical Studies, 115 (3):283-305. Papineau, D. 2003 The roots of Normativity, Oxford : Oxford University Press Peacocke, C. 1992 A Study of concepts , Cambridge, Mass: MIT Press 2004 The Realm of Reason , Oxford, Oxford University Press Pollock J and Cruz, J 1999 Contemporary theories of knowledge , Tatowa: Rowman and Littlefield Pritchard, D. 2007 'Recent Work on Epistemic Value', American Philosophical Quarterly 44 (2007), 85-110 Railton, P.2000 “Normative Force and Normative Freedom”, in Dancy 2000 and Railton 2003 2003 Facts Values and Norms , Cambridge: Cambridge University Press Ryan, S. 2003 “Doxastic Compatibilism and the Ethics of Belief,” Philosophical Studies, cxiv : 47–79. Scanlon, T 1998 What we owe to each other, Harvard: Harvard University Press Schröder , T. 2003 “ Davidson’s Theory of Mind is on normative”, Philosopher’s Imprint, 2, 1, http://hdl.handle.net/2027/spo.3521354.0003.001 Shah , N. 2003 “How Truth Regulates Belief”, The Philosophical Review, 112, 447-482 Shah, N & Velleman, D. 2005 “Doxastic Deliberation” , The Philosophical Review, 114(4):497-534 Skorupski J. 2010 The Domain of Reasons, Oxford, Oxford University Press

14 Sosa , E. 2007 A Virtue Epistemology , Oxford, Oxford University Press Steglish-Petersen A. 2006 “The Aim of Belief : no norm needed”, Philosophical Quarterly 56, 225, 500-516 Steup, M. 2000. “Doxastic Voluntarism and Epistemic Deontology” Acta Analytica 15: 25-56. Velleman , D. 2000 “The Aim of belief” in The Possibility of Practical Reason, Oxford: Oxford University Press Wedgwood , R. 2007 The Nature of Normativity, Oxford : Oxford University Press Weinberg, J. Nichols, S. Nichols, S. 2000 “Normativity and Epistemic Intuitions”, Philosophical Topics, 29, 1 & 2, 2001. Pp. 429-460. White , R “Epistemic permissiveness”, Philosophical Perspectives, 19:445–59. Williamson, T. 2000 Knowledge and its limits, Oxford : Oxford University Press Zagzebski, L 1996 Virtues of the Mind, Cambridge, Cambridge University Press